Artifact-centric business process models allow to describe artifacts (data objects) and their life cycles, which allow designers to focus on individual artifact in business processes, thus simplifies the design and analysis of business process model. However, this feature is a double-edged sword. The description of the relationships between artifacts becomes a new and nontrivial problem. It is better that the associations among business artifacts are specified at a high level as logical assertions. We think taking business constraints as complements of artifact-centric business operational model is an useful idea. Based on this consideration,in this paper, we propose an approach which combines both the declarative way and the procedural way in the construction of business processes. This flexibility can help designers to separate the parts of a business process that are more likely to change from those that are less likely to change. We propose a language TiLE to specify business constraints, and give complexity results on the satisfiability of TiLE. Moreover, we discussed how to enforce the constraints at run-time.

BPEL (Business Process Execution Language) has become the standard for specifying and executing workflow specifications for Web service composition invocation. A major weakness of BPEL is the lack of so-called "human workflow" support. The BPEL4People specification tries to amend this by adding human task support to BPEL. In this paper, we propose a formal model of BPEL4People using the CSP process algebra, and discuss some issues we found through analyzing the model. Although based on BPEL4People, this is a general work, and can also be viewed as a formal model of human workflow.

Web services choreography describes global models of service interactions among a set of participants. For an interaction to be executed, the participants must know the required channel(s) used in the interaction, otherwise the execution will get stuck. Because of dynamic composition, the initial channel set on each participant is often insufficient to meet the requirements. It is the responsibility of the participants to pass required channels owned (known) by one to some others. Since a choreography may involve many participants and complex channel constraints, it is hard for designers to specify channel passing in a choreography exactly as required. In this paper, we address the problem of checking whether a choreography lacks channels or has redundant channels, and how to automatically generate channel passing based on interaction flows of the choreography in the case of channel absence. Concretely, we propose a small language Chorcnamed for a channel interaction sub-language for modeling the channel passing aspect of choreography. Based on the formal operational semantics of Chorc, the algorithms for static checking choreography and generating channel passing are studied as well.

With the growth of interest on the web services, people pay increasinglyattention to the choreography, that is, to describe collaborations ofparticipants in accomplishing a common business goal from a globalviewpoint. In this paper, based on a simple choreography language and arole-oriented process language, we study some fundamental issues relatedto choreography, especially those related to implementation, includingsemantics, projection and natural projection, dominant role in choices anditerations, etc. We propose the concept of dominant role and somenovel languages structures related to it. The study reveals some cluesabout the language, the semantics, the specification and theimplementation of choreography.

We propose a language CDL as a formal model of simplified WS-CDL. The operational semantics of CDL is given, and static validation and verification of choreographies is studied. Some properties of the proposed model are verified using the SPIN model-checker, which illustrates the potential usage and benefits of the formal model

In the paper, we proposed a language called BPEL0 with its formal semantics as the foundations of WSBPEL. In this paper, we follow the way Van der Aalst proposed on pattern analysis in workflow languages (2003), and present the patterns for BPEL0. Moreover, the expressiveness of BPEL0 is also embodied by means of putting these patterns in the program environment composed of other programming operators. Those properties about the patterns with its environment are captured by the algebraic laws, which can be proven in the framework of BPEL0 semantic domain.